Because of the richness of the information available from NMR, it has often been argued that NMR is the most powerful analytical technique for molecular structure determination. However, NMR has been more successful with liquids or materials dissolved in solvents than with rigid solids. The basic problem in NMR of solids is that rapid molecular tumbling and diffusion are not naturally present to average out chemical shift anisotropy and dipolar couplings of abundant spin nuclides. Hence, the lines are normally broad and unresolved (often hundreds of ppm in width). A large number of techniques have been developed to improve the resolution in NMR of solids, but most modem techniques include extremely rapid spinning of the sample at the “Magic Angle” (the zero of the second Legendre polynomial, 54.7.degree.) with respect to B.sub.0. If the rotational rate is fast compared to chemical shift anisotropies and dipolar couplings (in units of Hz), the resolution is dramatically improved—often by two or three orders of magnitude. Even when the spinning is not fast enough to satisfy the above conditions, substantial improvements in resolution are generally obtained from the combination of MAS and multiple-pulse methods. Similar resolution problems are encountered in liquids in inhomogeneous systems, as in tissues and the mixtures of liquids and solids, because of susceptibility variations throughout the material. Here, relatively slow MAS is often effective in improving the spectral resolution of the liquid species by several orders of magnitude.
The progress in increasing sensitivity in NMR has been impressive over the past five decades—three to five orders of magnitude, depending on the application. The most significant, generally applicable contribution to increasing the signal to noise ratio, S/N or SNR, in the past decade has been the introduction of cryoprobes for homogeneous liquid samples, such as that by Marek, U.S. Pat. No. 6,677,751 B1, in which the receiver coil, critical tuning elements, and preamps are cryogenically cooled while the sample is kept at some experimentally desired temperature, usually near room temperature (RT). Using high-purity aluminum coils and single-layer capacitors near 25 K with the preamps perhaps at 80 K, the S/N may be increased on one or more channels in a multi-resonant probe by typically a factor of three to four.
Most modern NMR applications are directed at structure determinations of complex macromolecules, where it is often desirable to utilize a probe with high S/N at two or three different frequencies simultaneously and perhaps additionally be able to lock the field on the .sup.2H resonance. RF circuit efficiencies in 3 to 5 mm triple-resonance MAS probes with a single rf solenoid for signal reception at very high fields are typically in the range of 25 35% at the low-frequency (LF) and 15 40% at the mid-frequency (MF). For examples of a triple-resonance MAS circuit with a single sample solenoid, see my U.S. Pat. No. 5,424,645 or the circuit by Martin, Paulson, and Zilm in “Design of a Triple Resonance MAS probe for High Field Solid-state NMR,” in Rev. Sci. Instrum., 74, 6, 3045 3061, 2003. Note that rf efficiency is defined as the percent of rf transmit power dissipated in the sample and the sample coil, as in principle other losses can be eliminated.
Significantly higher rf efficiencies on all channels in MAS have been achieved using a cross-coil for .sup.1H and a solenoid for the MF and LF, as in my U.S. Pat. No. 6,130,537 or as discussed by Doty et al in “Magnetism in NMR Probe Design Part II: HR MAS,” in Concepts in Magn. Reson., Vol 10(4), 239 260, 1998. Still efficiencies are generally in the range of 30 50% for both the LF and the MF channels. Moreover, most advanced NMR MAS applications are now at 11.7 T or higher and also are requiring magic angle gradients, as disclosed by Cory in U.S. Pat. No. 5,872,452, and automatic sample change, all of which tend to push rf efficiencies toward the low ends of the above ranges.
Thus far, high-resolution (HR) NMR probes in which the sample coil and other circuit elements are at cryogenic temperatures have only been demonstrated for liquid samples in which the sample tube is aligned with the polarizing field, B.sub.0. Triple-resonance MAS probes in which both the sample and the sample coil may be simultaneously cooled to essentially the same temperature, both below 120 K, have been commercially available for high field magnets with 40 mm and larger RT shim bores for at least 15 years. In a few cases, cooling of the sample and sample coil to as low as 30 K has been possible in double-resonance MAS probes, but most critical tuning elements in such have not usually been cooled.
Unfortunately, cryogenic cooling of the sample coil in an MAS probe in which the sample is not also near the cryogenic temperature of the sample coil appears impractical within the space constraints of normally available “standard bore” or “narrow bore” high-field NMR magnets, where the bore of the RT shim tube system is typically 40 mm. Such a probe also appears impractical even in “mid-bore” magnets, where this dimension is typically 45 mm or 51 mm. An HR CryoMAS probe, capable of triple-resonance MAS NMR in which the sample coils are cryogenically cooled while the sample is at RT, does appear practical in magnets with RT shim bores greater than 60 mm, and such will be the subject of another patent application by Doty.
Using the same coil for both transmit and receive has been the preferred approach in NMR spectroscopy probes, both for liquids and solids, for at least three decades. In this case, Hoult's principle of reciprocity, which at least in its popular usage states that the NMR S/N during reception is, among other things, proportional to the square root of the circuit efficiency for generating a transverse rf magnetic field within the sample during transmit, has seldom been challenged. Not surprisingly, maximizing rf circuit efficiency in multi-tuned NMR probes has been a major focus of several international firms over the past two decades.
While it is not difficult to achieve rf efficiency above 85% in single-tuned circuits, much lower efficiencies are always obtained in double- or triple-tuned circuits, especially for MAS probes, as noted above. The challenges are greater in MAS probes than in liquids probes primarily because the circuits must also be designed to handle very high power, which requires larger circuit components, and because the spinner assembly interferes with efficient lead routing, especially when automatic sample change is also desired in narrow-bore magnets.
Reciprocity, as defined above, fails to be valid when the various loss mechanisms (sample, sample coil, capacitors, shields, etc.) are at significantly different temperatures, as the transmit efficiencies are determined by the various resistances in the circuit, but the noise power during receive is proportional to both the resistance and its temperature. One example where reciprocity fails is in cryoprobes, such as that disclosed in U.S. Pat. No. 5,508,613, where the sample and perhaps some other minor loss components are much warmer than the sample coil.
In conventional liquids NMR probes, the noise is dominated by that from the sample coil, and cooling it is the primary objective in current cryoprobes for liquids, though of course attention is also paid to reducing the noise from other circuit tuning elements and the preamp. In triple-resonance MAS, on the other hand, the most significant single noise source can be a secondary tuning coil, and the total noise from the (critical) high-power tuning capacitors may be similar to that from the sample coil. This invention addresses MAS probes where cryogenic cooling of the sample coil is impractical. And in these cases, it allows the noise from the secondary tuning coil to be reduced by a factor of six (both its resistance and temperature are reduced), and that from the high-power capacitors to be reduced by a factor of three (often their resistance has little temperature dependence), assuming these reactive elements are cooled to 100 K, for example. Moreover, a circuit is disclosed that allows the noise contributions from the variable capacitors to be reduced to a few percent of the total. It turns out that the inventive circuit is also advantageous even when the cold zone is not cooled, but its advantages become substantial upon cooling.
The objective of this invention is to permit substantial improvement in S/N in triple-resonance HR MAS NMR in high field magnets without cooling the sample, especially where the RT shim bore is less than 55 mm. The low LF and MF efficiencies in such MAS probes suggest there is considerable opportunity for noise reduction, at least on the MF and LF channels, by cooling critical circuit elements other than the sample coil and thereby reducing their contribution to circuit noise power by a factor of three to six.